import numpy as np
import matplotlib
import matplotlib.pyplot as plt
import librosa
import soundfile as sf
from PIL import Image
import cv2
from IPython.display import Audio, HTML, display
import os

%matplotlib inline

print("Library versions:")
print(f"NumPy: {np.__version__}")
print(f"Matplotlib: {matplotlib.__version__}")
print(f"Librosa: {librosa.__version__}")
print(f"OpenCV: {cv2.__version__}")

from IPython.display import Audio
from pathlib import Path
import librosa.display as lbd

Library versions:
NumPy: 2.2.6
Matplotlib: 3.10.6
Librosa: 0.11.0
OpenCV: 4.12.0

path = Path("data/audio_proklamasi_soekarno.wav")
assert path.exists(), f"File tidak ditemukan: {path.resolve()}"

with sf.SoundFile(path) as f:
    sr = f.samplerate
    channels = f.channels
    frames = len(f)

duration = frames / sr
label = "mono" if channels == 1 else "stereo"

print(f"Sample rate (Hz): {sr}")
print(f"Durasi (detik): {duration:.2f}")
print(f"Jumlah kanal: {channels} ({label})")
print(f"Jumlah total sampel: {frames}")
Audio(path)

Sample rate (Hz): 44100
Durasi (detik): 48.62
Jumlah kanal: 2 (stereo)
Jumlah total sampel: 2144256

y, sr = sf.read(path, dtype="float32", always_2d=True)
n, ch = y.shape

if ch == 1:
    stereo = np.repeat(y, 2, axis=1)
    info = "File asli mono, dibuat dual-mono untuk visual stereo"
else:
    stereo = y
    info = "File asli stereo"

mono_vec = stereo.mean(axis=1)
mono = mono_vec[:, None]

# Vektor waktu
t = np.arange(n) / sr

print("=== Info matriks ===")
print(f"Stereo shape: {stereo.shape}  (kolom 0=L, 1=R)")
print(f"Mono shape  : {mono.shape}    (kolom tunggal)")
print(info)

# 1) Mono
plt.figure(figsize=(12, 2.6))
plt.plot(t, mono_vec)
plt.title("Waveform Mono")
plt.xlabel("Waktu (detik)")
plt.ylabel("Amplitudo")
plt.xlim(0, t[-1])
plt.tight_layout()
plt.show()

# 2) Stereo 1 graf (overlay L dan R)
plt.figure(figsize=(12, 2.6))
plt.plot(t, stereo[:, 0], label="Kiri")
plt.plot(t, stereo[:, 1], label="Kanan")
plt.title("Waveform Stereo - overlay")
plt.xlabel("Waktu (detik)")
plt.ylabel("Amplitudo")
plt.legend(loc="upper right")
plt.xlim(0, t[-1])
plt.tight_layout()
plt.show()

# 3) Stereo kanal kiri
plt.figure(figsize=(12, 2.6))
plt.plot(t, stereo[:, 0])
plt.title("Waveform Stereo - kanal kiri")
plt.xlabel("Waktu (detik)")
plt.ylabel("Amplitudo")
plt.xlim(0, t[-1])
plt.tight_layout()
plt.show()

# 4) Stereo kanal kanan
plt.figure(figsize=(12, 2.6))
plt.plot(t, stereo[:, 1])
plt.title("Waveform Stereo - kanal kanan")
plt.xlabel("Waktu (detik)")
plt.ylabel("Amplitudo")
plt.xlim(0, t[-1])
plt.tight_layout()
plt.show()

=== Info matriks ===
Stereo shape: (2144256, 2)  (kolom 0=L, 1=R)
Mono shape  : (2144256, 1)    (kolom tunggal)
File asli stereo

y_mono = y.mean(axis=1)

# Parameter STFT
n_fft = 1024
hop_length = 256
window = "hann"

# STFT -> magnitude -> log-dB
D = librosa.stft(y=y_mono, n_fft=n_fft, hop_length=hop_length, window=window, center=True)
S_db = librosa.amplitude_to_db(np.abs(D), ref=np.max, top_db=80)

# Plot
plt.figure(figsize=(12, 4))
img = lbd.specshow(S_db, sr=sr, hop_length=hop_length, x_axis="time", y_axis="linear")
cbar = plt.colorbar(img, format="%+2.0f dB")
cbar.set_label("Level (dB)")
plt.title("Spectrogram log-dB (mono)")
plt.xlabel("Waktu (detik)")
plt.ylabel("Frekuensi (Hz)")
plt.tight_layout()
plt.show()

n_mfcc = 13
n_mels = 40

# Mel-spectrogram -> MFCC
S_mel = librosa.feature.melspectrogram(
    y=y_mono, sr=sr, n_fft=n_fft, hop_length=hop_length, n_mels=n_mels, power=2.0
)
mfcc = librosa.feature.mfcc(S=librosa.power_to_db(S_mel, ref=np.max), n_mfcc=n_mfcc)

# Plot heatmap MFCC
plt.figure(figsize=(12, 4))
img = lbd.specshow(
    mfcc,
    x_axis="time",
    sr=sr,
    hop_length=hop_length,
)
cbar = plt.colorbar(img)
cbar.set_label("Koefisien MFCC")
plt.title("MFCC 13 koefisien")
plt.ylabel("Koefisien (1 sampai 13)")
plt.xlabel("Waktu (detik) - unit frame")
plt.tight_layout()
plt.show()

path = Path("data/gambar_kobe_bryan.jpeg")

img_bgr = cv2.imread(str(path), -1)
if img_bgr is None:
    raise RuntimeError("Gagal memuat gambar")

if img_bgr.ndim == 2:
    img_rgb = img_bgr
elif img_bgr.shape[2] == 3:
    img_rgb = cv2.cvtColor(img_bgr, cv2.COLOR_BGR2RGB)
elif img_bgr.shape[2] == 4:
    img_rgb = cv2.cvtColor(img_bgr, cv2.COLOR_BGRA2RGBA)
else:
    raise ValueError(f"Bentuk kanal tidak didukung: {img_bgr.shape}")

plt.figure(figsize=(8, 5))
plt.imshow(img_rgb)
plt.title("Tampilan benar - RGB")
plt.axis("off")
plt.tight_layout()
plt.show()

plt.figure(figsize=(8, 5))
plt.imshow(img_bgr)
plt.title("Tampilan salah - BGR")
plt.axis("off")
plt.tight_layout()
plt.show()

img = Image.open(path)
mode = img.mode
W, H = img.size

arr = np.asarray(img)
dtype = arr.dtype
if arr.ndim == 2:
    channels = 1
else:
    channels = arr.shape[2]

mem_bytes = arr.nbytes
mem_mb = mem_bytes / (1024**2)

print("=== Informasi Dasar Gambar ===")
print(f"Path                 : {path.name}")
print(f"Dimensi (H x W)      : {H} x {W} piksel")
print(f"Jumlah kanal         : {channels}")
print(f"Tipe data (dtype)    : {dtype}")
print(f"Mode warna (PIL)     : {mode}")
print(f"Ukuran di memori     : {mem_mb:.2f} MB  ({mem_bytes} bytes)")

=== Informasi Dasar Gambar ===
Path                 : gambar_kobe_bryan.jpeg
Dimensi (H x W)      : 876 x 736 piksel
Jumlah kanal         : 3
Tipe data (dtype)    : uint8
Mode warna (PIL)     : RGB
Ukuran di memori     : 1.84 MB  (1934208 bytes)

R = img_rgb[:, :, 0].ravel()
G = img_rgb[:, :, 1].ravel()
B = img_rgb[:, :, 2].ravel()

bins = 256
hist_R, edges = np.histogram(R, bins=bins, range=(0, 255))
hist_G, _     = np.histogram(G, bins=bins, range=(0, 255))
hist_B, _     = np.histogram(B, bins=bins, range=(0, 255))
x = edges[:-1] + 0.5  

def smooth(y, k=7):
    k = max(1, int(k))
    if k % 2 == 0: k += 1
    kernel = np.ones(k) / k
    return np.convolve(y, kernel, mode="same")

hist_R_s = smooth(hist_R, k=7)
hist_G_s = smooth(hist_G, k=7)
hist_B_s = smooth(hist_B, k=7)

plt.figure(figsize=(8, 5))
plt.imshow(img_rgb)
plt.axis("off")
plt.tight_layout()
plt.show()

plt.figure(figsize=(10, 5))
plt.plot(x, hist_R_s, label="Red", color="red", linewidth=2)
plt.plot(x, hist_G_s, label="Green", color="green", linewidth=2)
plt.plot(x, hist_B_s, label="Blue", color="blue", linewidth=2)
plt.title("Histogram Warna Gambar (RGB)")
plt.xlabel("Intensitas Piksel (0-255)")
plt.ylabel("Frekuensi")
plt.xlim(0, 255)
plt.legend()
plt.tight_layout()
plt.show()

video_path = Path("data/video_batle_anime_demon_slayer.mp4")
assert video_path.exists(), f"File tidak ditemukan: {video_path.resolve()}"

cap = cv2.VideoCapture(str(video_path))
if not cap.isOpened():
    raise RuntimeError("Gagal membuka file video")

width  = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH)  or 0)
height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT) or 0)
fps    = float(cap.get(cv2.CAP_PROP_FPS) or 0.0)
total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT) or 0)

if fps <= 0 and total_frames > 1:
    ok, _ = cap.read()
    ok2, _ = cap.read()
    if ok and ok2:
        t_sec = cap.get(cv2.CAP_PROP_POS_MSEC) / 1000.0
        if t_sec > 0:
            fps = 1.0 / t_sec
    cap.set(cv2.CAP_PROP_POS_FRAMES, 0)

duration = (total_frames / fps) if fps > 0 else np.nan

if width >= 3840 and height >= 2160:
    resolution_class = "4K (Ultra HD)"
elif width >= 2560 and height >= 1440:
    resolution_class = "QHD (1440p)"
elif width >= 1920 and height >= 1080:
    resolution_class = "Full HD (1080p)"
elif width >= 1280 and height >= 720:
    resolution_class = "HD (720p)"
elif width >= 640 and height >= 480:
    resolution_class = "SD (480p)"
else:
    resolution_class = "Low Resolution"

print("=== Metadata Video ===")
print(f"Resolusi       : {width} x {height}")
print(f"Frame Rate     : {fps:.2f} fps")
print(f"Total Frame    : {total_frames}")
print(f"Durasi         : {duration:.2f} detik" if fps > 0 else "Durasi         : tidak dapat dihitung (FPS=0)")
print(f"Klasifikasi    : {resolution_class}")
cap.release()

=== Metadata Video ===
Resolusi       : 1280 x 720
Frame Rate     : 24.00 fps
Total Frame    : 559
Durasi         : 23.29 detik
Klasifikasi    : HD (720p)

cap = cv2.VideoCapture(str(video_path))
assert cap.isOpened(), "Gagal buka video"

total = int(cap.get(cv2.CAP_PROP_FRAME_COUNT) or 0)
fps   = float(cap.get(cv2.CAP_PROP_FPS) or 24.0)
dur   = total / fps if fps > 0 else 0.0

def read_rgb_at_index(cap, idx, backoff=5):
    for b in range(backoff+1):
        i = max(0, idx - b)
        cap.set(cv2.CAP_PROP_POS_FRAMES, i)
        ok, bgr = cap.read()
        if ok and bgr is not None:
            rgb = cv2.cvtColor(bgr, cv2.COLOR_BGR2RGB)
            t = i / fps
            return rgb, i, t
    return None, None, None

def read_rgb_at_time(cap, t_sec):
    cap.set(cv2.CAP_PROP_POS_MSEC, max(0, 1000.0 * t_sec))
    ok, bgr = cap.read()
    if not ok or bgr is None:
        return None, None, None
    pos = int(cap.get(cv2.CAP_PROP_POS_FRAMES))
    rgb = cv2.cvtColor(bgr, cv2.COLOR_BGR2RGB)
    return rgb, pos, pos / fps

i0 = 0
i1 = max(0, total // 2)
i2 = max(0, total - 1)

f0 = read_rgb_at_index(cap, i0)[0]
f1 = read_rgb_at_index(cap, i1)[0]

f2, i2_real, t2_real = read_rgb_at_index(cap, i2, backoff=10)
if f2 is None:
    f2, i2_real, t2_real = read_rgb_at_time(cap, dur - 1.0 / fps - 1e-3)

cap.release()

titles = [
    f"Frame 0  t=0.00s",
    f"Frame {i1}  t={i1/fps:.2f}s",
    f"Frame {i2_real if i2_real is not None else 'N/A'}  t={t2_real:.2f}s" if t2_real is not None else "akhir (gagal)"
]

plt.figure(figsize=(12,4))
for k, frame in enumerate([f0, f1, f2]):
    ax = plt.subplot(1,3,k+1)
    if frame is None:
        ax.text(0.5, 0.5, "gagal baca", ha="center", va="center")
        ax.axis("off")
    else:
        ax.imshow(frame)
        ax.axis("off")
    ax.set_title(titles[k])
plt.suptitle("Awal  Tengah  Akhir  (RGB)")
plt.tight_layout()
plt.show()

Tugas Analisis Multimedia: Audio, Gambar, Video¶

Deskripsi Tugas¶

Checklist Kelengkapan (Centang ✅ saat selesai)¶

🎵 Bagian Audio¶

🖼️ Bagian Gambar¶

📹 Bagian Video¶

📝 Analisis & Dokumentasi¶

Bagian A — Audio¶

A1. Deskripsi Data¶

A2. TODO: Muat & Metadata¶

A3. TODO: Waveform¶

A4. TODO: Spectrogram log-dB¶

A5. TODO: MFCC¶

A6. Analisis Ringkas (Wajib)¶

Bagian B — Gambar¶

B1. Deskripsi Data¶

B2. TODO: Baca & Tampilkan (RGB)¶

B3. TODO: Informasi Dasar¶

B4. TODO: Histogram Warna¶

B5. Analisis Ringkas (Wajib)¶

Bagian C — Video¶

C1. Deskripsi Data¶

C2. TODO: Baca & Metadata¶

C3. TODO: Tampilkan 3 Frame (Awal–Tengah–Akhir)¶

C4. Analisis Ringkas (Wajib)¶

Perbandingan & Kesimpulan¶

Perbandingan Representasi Media¶

Audio (1D - Temporal)¶

Gambar (2D - Spasial)¶

Video (2D + Waktu - Spatio-temporal)¶

Refleksi Pembelajaran¶

3 Poin yang Saya Pelajari:¶

2 Hal yang Masih Membingungkan/Ingin Diperdalam:¶

Sumber Data & Referensi¶

Panduan Pengumpulan¶

📁 Berkas yang Harus Dikumpulkan¶

Wajib:¶

📅 Informasi Pengumpulan¶

✅ Checklist Sebelum Submit¶

Export ke PDF:¶